Basin Planning for Ganga River Basin in India

(1)

Conceptualization of River Basin Model, Surface water - Ground water Interaction Analysis, and Environmental Flow Assessment

2 December 2016

(2)

(3)

Basin Planning for Ganga River Basin in India

Conceptualization of River Basin Model,

Surface water Ground water Interaction Analysis, and Environmental Flow Assessment

1220123-000

Marnix van der Vat, Arthur Lutz, Mark Hegnauer, Pascal Boderie , Frans Roelofsen, Fernando Magdaleno Mas, Victor Langenberg and Kees Bons (editor) with contributions from all team members

(4)

(5)

Keywords

India, Ganga River, water quality, ecology, water resources, hydrology, geohydrology, information system, GIS, water demand, irrigation, environmental flows, collaborative modeling

Summary

Part A of this report describes the conceptualization and set-up of the River Basin Model for Strategic Planning of the Ganga Basin. It elaborates further on the information provided in the Terms of Reference, the proposal and the inception report. Moreover, this report is informed by the first results of the stakeholder involvement process and the data collection efforts. The conceptualization of the model is not expected to be further modified. However, the set-up of the River Basin Model as described here, will serve as a starting point for the collaborative modeling phase. During this phase, the model set-up will be fine-tuned to meet the requirements of the stakeholders and to incorporate their knowledge and understanding of the functioning of the system.

Part B describes in how the surface water (SW) and groundwater (GW) relate and gives an example of the interaction in the Ganga basin. Furthermore it describes the setup of the SW- GW assessment that will be carried out

Part C describes how a multi-scale environmental flow assessment will be developed along the Ganga River course, applying a consultative process based on sound scientific analyses.

To allow informed decision-making on the sustainable use of the Ganga River system, it is important to know the consequences any changes in use may have on the ecohydrological functioning of the Ganga river system and the types of ecosystem services (ESS) this system offers to society. Our approach consists of four steps:

a) Review of earlier and on-going environmental flows work and partnerships for the Ganga River.

b) Basin-wide assessment of hydrological alteration of the Ganga River system.

c) Identification of flow-ecology-ESS relationships based on expert judgment.

d) Incorporation of information on key relationships collected into modeling framework and dashboard.

This report (together with the progress report) contributes to Project Milestone 3 and combines the following deliverables as described in the Inception report into one report:

Deliverable 5 Report describing model conceptualization and setup, and Deliverable 6 Report with detailed approach for Task 2: Surface-groundwater Interaction Analyses and for Task 3:

Environmental Flow Assessments.

State

Final

(6)

(7)

Introduction 1

The Ganges is the most populated river basin in the world and is home to half the population of India including two-thirds of the nation’s poor people. The basin provides over one-third of the available surface water in India and is the focus of over half the national water use – 90 percent of this being in irrigation.

The ecological health of the Ganga River and some if its tributaries has deteriorated significantly as a result of high pollution loads (from point and non-point sources), high levels of water abstraction for consumptive use (mostly for irrigation but also for municipal and industrial uses), and other flow regime and river modifications caused by water resources infrastructure (dams and barrages for diverting and regulating the river and generating hydropower).

The Government of India has committed to an ambitious goal of rejuvenating the Ganga and is committing significant funds to address the problem. However, in addition to the technical complexity and scale, Ganga rejuvenation is an inherently “wicked problem” given the wide diversity of stakeholder values and perspectives and the political and institutional dimensions that come from distributed responsibilities across multiple jurisdictions and institutions.

The World Bank has assigned Deltares and its partners AECOM India and FutureWater to carry out the project ”Analytical Work and Technical Assistance to support Strategic Basin Planning for Ganga River Basin in India”.

As outlined in the Terms of Reference and our proposal, the key objectives of the project are:

(i) Significantly strengthen the capability of relevant central and state government agencies to undertake comprehensive evidence-based strategic basin planning for the Ganga River basin

(ii) Develop, document and disseminate (through detailed analytical work and stakeholder engagement) a set of plausible scenarios that balance significantly improving the health of the river and maintaining an acceptable level of economic productivity;

(iii) Build stronger and more accessible information and knowledge base to guide on- going dialogue around and management of the Ganga River basin; and

(iv) Establish on-going multi-stakeholder engagement processes in the basin to support strategic basin planning.

These objectives will be achieved by:

(i) Developing a detailed and robust water resources planning model for the entire Ganga basin in India and training central and state government engineers and planners in its use;

(ii) Characterizing and analyzing surface-groundwater interactions across the basin using this information to refine the river modeling;

(iii) Undertaking a multi-scale environmental flow assessment across the basin and using these assessments to inform the scenario modeling;

(iv) Developing, modeling and disseminating a series of plausible scenarios that explore alternative options for improving water management including improving river health;

(10)

(v) Establishing and facilitating a multi-stakeholder consultation process (inside and outside of government) to guide and share the work above; and

(vi) Ensuring wide access to the models and analyses and quality documentation of these.

This report describes the conceptualization and set-up of the River Basin Model for Strategic Planning of the Ganga Basin. It elaborates further on the information provided in the Terms of Reference, the proposal and the inception report. Moreover, this report is informed by the first results of the stakeholder involvement process and the data collection efforts. The conceptualization of the model is not expected to be further modified. However, the set-up of the River Basin Model as described here, will serve as a starting point for the collaborative modeling phase. During this phase, the model set-up will be fine-tuned to meet the requirements of the stakeholders and to incorporate their knowledge and understanding of the functioning of the system.

The River Basin Model describes the functioning of the water system of the Ganga Basin within India with respect to rainfall-runoff, flow storage and diversion, water use and water quality and ecology. The interaction between surface and groundwater is included in the model concept, but is described in a separate report. The aim of the model is to analyses the impact of possible future developments, such as climate change and socio-economic scenarios and the possible management strategies. The impact will be presented in the form of values for indicators, which that will be determined together with the stakeholders.

The report continues with the elaboration of how the Surface Water-Ground Water interaction is assessed in the project and a description of the Environmental Flow assessment. These assessments are based on and will make use of the model framework.

This report contributes to Project Milestone 3 and combines the following deliverables into one report:

· Deliverable 5 Report describing model conceptualization and setup

· Deliverable 6a Report with detailed approach for Task 2: Surface-groundwater Interaction Analyses and

· Deliverable 6b Report with detailed approach for Task 3: Environmental Flow Assessments

(11)

A - RIVER BASIN MODEL CONCEPTUALIZATION Context: The collaborative modeling process 2

The collaborative modeling process commenced during the inception phase with the meeting of stakeholders at different basin-wide and state-level meetings and workshops (see also the inception report). At these meetings the stakeholder responses to the project and its set-up were solicited, also in addition to their initial ideas on the most important water-related issues confronting the basin.

These initial ideas were subsequently complemented with information received from questionnaires that were sent to a wide range of state-level stakeholder organizations.

Several questions in this questionnaire related to the perceived issues, their impacts and their causes.

Based on the input received through the meetings, workshops and questionnaires, the project team identified issues that play a role in the different states. A 1-day basin-wide workshop with key stakeholders from both the central level as well as each of the 11 Ganga states is planned in July to provide all participants feedback on this period and insight in other stakeholder’s responses. This meeting will be used to further validate basin-wide issues with the addition of state perspectives; thereby integrating an even larger number of perspectives into the basin-wide assessment and laying the foundations for later inter-state cooperation regarding the shared use of the basin’s water resources. A particular focus of this workshop will also be to define the initial set of indicators to be used by decision makers in the basin for water resources planning. The final set of indicators will be finalized over the remainder of the project and will feature in the dashboard to be developed during later stages of the project.

To validate and further elaborate these findings for input into the technical modeling process, another series of workshops will be organized in the period July-October 2016:

A. A 1-day basin-wide workshop with specialists from central level agencies. This will be used to validate basin-level issues and their causes, and to further develop and reach agreement regarding the ensuing modeling process. Involving technical specialists from relevant central level agencies will provide a solid opportunity to consider the more technical issues in an integrated manner, share stakeholder perspectives, build inter-agency cooperation, and generate interest in the modeling activities to come.

B. A series of 2-days workshop in each of the 11 Ganga states with relevant state level stakeholders. These will be used to validate water-related issues at the state level with their causes, impacts, possible mitigation measures and indicators. In doing so, participants will gain an integrated shared understanding of each state’s water resources system in addition to an appreciation of the priorities and perspectives of the different stakeholders. The causal chains will then be translated as much as possible to maps as the first step in the schematization of the different models. This series of workshops will contain the following elements:

(12)

1. Issues validation: The issues identified during the previous meetings and the questionnaires will be discussed with the participants, giving them the opportunity to

collectively prioritize these and add and/or subtract additional issues or

measures to this list. Post-it stickers will be used to in this exercise as per the example illustrated in the figure on the right.

2. Identify the causal chains:

The agreed set of issues from the previous step will then be discussed in

greater detail, to identify those that can be connected to each other in causal relationships. Both the root causes and impacts for each issue will be identified and their qualitative relationships established. In addition, any potential measures that have been identified will be assessed according to their influence on both their targeted and interrelated factors. Potential indicators to measure these impacts will also be discussed and defined for later use in the dashboard. It is anticipated that during the process of linking the different factors together in causal chains, additional new factors (causes of causes, impacts of impacts) will most likely be identified. An example of a possible result is given in the picture below:

Figure 2-1: Issues on stickers for group discussion

Figure 2-2: Causal chains emerging from group sessions

(13)

3. Translation of the causal chains to features on the map. The causal chains will then be analyzed for any physical features, or any issues that can be assigned to particular locations. These will then be drawn onto maps, which will result in a preliminary schematization of the physical system that can be used to improve the schematization for the different models. An example of how such a map might look is given in the picture below.

In the wake of these workshops, the modeling team will use the additional information gleaned and decisions taken to further develop the (geo-)hydrological, water resources, and water quality and ecological models, which will then be collaboratively validated and used for the assessment of impacts of packages of measures in later phases of the project.

To improve communication during the collaborative modeling workshops we use the River flow regime-ESS concept (Figure 2-4).

Figure 2-3: Map showing participant's inputs

(14)

Figure 2-4: River Ganga Flow regimes- Ecosystem Services concept

The flow regime-ESS concept is an ecosystem-based and dynamic approach which fits the dynamic, non-linear nature of social/ecological systems and has proven to be beneficial in improving the processes by offering an easier ‘language’ to communicate stakeholder’s positions and interests. The concept calls for the joint development of system understanding of the functional inter-relationships between the Ganga River and social system, the basis for river basin management.

For an effective application of the Collaborative Modeling approach, a content-related analytical framework and an adapted participatory process of involving stakeholders are essential. Whereas a common understanding of the value of the goods and services that the healthy Ganga river basin ecosystem can provide, and the diminution of these values by our actions, is the key to a better approach to Ganga river basin management.

The collaborative modeling seeks to:

• Discover common interests in the state and health of the land-water system of the River Ganga.

• Disclose expertise and current understanding of the flow regime and important ESS.

• Help to classify and narrow down on most important ESS

• Establish relationships between flow regime changes and key ESS.

• Identify important region in the Ganga zonation where a given ESS is expected to be most emphasized or realized along a river continuum.

During the workshops, the following set of questions will be discussed to gain knowledge about the flow-ecology-ESS interactions along the river system:

1. Where are the principal locations / tributaries for most important goods and benefits of the River Ganga?

a. Can you shortlist very important goods and benefits?

b. Can you establish links with the current status at locations / tributaries?

2. What are the requirements for these goods and services in terms of a certain river discharge?

(15)

3. What are the requirements for certain goods and services in terms of water quality?

4. Are there any specific sites to be targeted for species protection (e.g. dolphin or reptile habitats)?

5. Are there any key locations where floodplains, wetlands or whole tributaries have altered or disappeared?

6. Are there any key locations where goods and services losses are evident?

a. Are the main causes for losses known or can reasons be derived?

7. Where are the key locations where environmental pollution is evident?

a. What characterizes the remaining flora and fauna?

b. Can trends be established?

8. Where are the key locations where water is heavily polluted (e.g. with toxic chemicals, organic wastes, nutrients?

9. Are there any existing plans to supply water or improve its quality (e.g. by building STPs) for goods and services in the future?

a. Where?

b. When?

c. How much water?

d. What WQ measures are to be taken?

(16)

Components of the River Basin Model and their interaction 3

The River Basin Model consists of several components that interact with each other (Figure 3-1). The model as described in part A of this report deals with:

· The hydrology (Chapter 4);

· The groundwater model (Chapter 5);

· The water resources (Chapter 6);

· The water quality (Chapter 7);

The application of the integrated modeling framework is described (Chapter 8) as well as the assessment of impacts on ecology and ecosystem services and the determination of environmental flow regimes that make use of the models (Chapters 10-12).

A separate report is prepared on the storage of model input and output in the GangaWIS (Water Information System) and the presentation of results on a dashboard.

Figure 3-1 Components of the River Basin Model and their interaction

The description of the hydrology and the rainfall-runoff process has been divided over two different models: SPHY and WFlow. They are both fully distributed models working on a grid of square cells. SPHY is used to describe the hydrological process in the mountainous areas in the Himalaya. This model has been selected, because it is specifically designed for glacier and snow hydrology and because it has been previously applied successfully for the Himalayas. Section 4.2 provides a detailed description of the concepts, set-up, data requirements and calibration of the SPHY model.

The rainfall-runoff processes for the non-mountainous part of the Ganga Basin are described by the WFlow model. This is a general purpose hydrological model that also allows calculation of water levels and contains a simplified module to describe flooding in the flood plains of the river. The river discharges calculated by the SPHY model for the Himalayas are used as upstream boundaries for the WFlow model. The information on discharges and water levels calculated by WFlow are used by the groundwater model to describe the interaction

(17)

between surface and groundwater. This information can be used again as input for a next run of WFlow. In this way an iteration process is created to ensure consistent results from both models (see further the report on the detailed approach for surface – groundwater interaction). The application of the WFlow model is described in section 4.3.

The water resources model RIBASIM describes the management and use of water. Its hydrological input is derived from the river discharges calculated by WFlow. RIBASIM uses a schematization of links and nodes to describe the flow of water in the rivers, the storage in reservoirs, the diversion into canals and the use and return flow by different functions. Water can be used from rivers and canals or from groundwater. Conjunctive use of surface and groundwater is also possible. Furthermore, return flows can be divided over rivers, canals and groundwater. This an important aspect for the description of the water system in the plains of the Ganga Basin, where extensive leakage from irrigation canals, feeds the groundwater aquifers, that are themselves used for irrigation water supply. Therefore, the RIBASIM model is also linked to the groundwater model by prescribing extraction and infiltration rates. The concept, set-up and data requirements of the RIBASIM model are described in Chapter 4 together with a description of the joint calibration of WFlow and RIBASIM.

Water quality can be assessed from the results of the RIBASIM model by tracing the origin of water to different sources of pollution. This allows for a risk assessment of water quality problems. For the most important pollutants for which enough data become available, the RIBASIM results will be combined with a pollutant load estimation to model the water quality with DWAQ. The DWAQ model is described in Chapter 6.

The impact on the ecology and ecosystem services of the results of the models presented above with respect to discharges, water levels and water quality will be evaluated using knowledge rules. These rules are site specific and will be developed during the project together with the stakeholders. A further description of this component can be found in the report detailing the approach for the environmental flow assessment. A description of the links with the other models is provided in Chapter 6.

All model input and all relevant output will be stored in the GangaWIS. The exchange of information between the components of the River Basin Model will also take place through the GangaWIS. The management of different versions of model input and output, to represent different scenarios and strategies, will be included in the GangaWIS. Furthermore, the model results stored in the GangaWIS will provide the input for the presentation of results in the dashboard. A separate report will be prepared describing the design of GangaWIS.

Most of the components of the River Basin Model are open source. This applies to SPHY, WFlow, iMOD, MODFLOW and DWAQ. This means that both the source code and the executable form of the software is publicly available on internet to all interested parties and can be downloaded free of charge. The RIBASIM software is licensed software under transition to become open source. Deltares as the owner of RIBASIM has agreed to make the software available in an executable form free of charge for application with India during and after execution of this project. The GangaWIS and the evaluator for the knowledge rules are built entirely with open source components. Any new code prepared for this whole system during this project will be made available to all interested parties free of charge.

The project area is defined in the terms of reference to “encompass the entire Ganga River basin in India including all tributaries upstream of Farakka Barrage on the Ganga River”.

(18)

that enter the Ganga via the Nepalese tributaries”. Therefore, the combined application of the hydrological models SPHY and WFlow will cover the entire Ganga Basin upstream of Farakka Barrage including the parts of the upstream basin located in Nepal and China. This allows for the requested robust assessment of the upstream flows. The application of the other models will be mostly limited to the Indian part of the Ganga Basin upstream of Farakka Barrage, with the possible exception of the major reservoirs on the Nepalese tributaries that might have to be included in RIBASIM to describe consistently their operation.

The initial set-up of the models SPHY and WFlow (and also iMOD/MODFLOW) is on a cell size of 1x1km. During project execution, the grid size might be enlarged to reduce computation times, but only if this does not compromise unacceptably the accuracy of the results. The models RIBASIM and DWAQ work on a schematization of the river basin as links and nodes. A preliminary set-up of these is presented in this report, but this schematization will be further fine-tuned during the collaborative modeling together with the stakeholders.

The models will be applied for different periods. For calibration this depends on the length of the time series available for model input and for comparison of model results with measurements. For the discharges used to calibrate SPHY, WFlow and RIBASIM (within India) it is foreseen that this period will cover 30 years from 1985 to 2015. For the water quality the period foreseen is 2001 to 2015. The time step of the calculations in SPHY and WFlow will be one day and for RIBASIM and DWAQ one month. The reason for this is that the main hydrological processes take place within periods of days and require calibration on this temporal resolution. The main processes regarding water resources and water quality, on the other hand, can be dealt with on the larger time scale of a month.

The aim of the River Basin Model is to support strategic planning on a basin level. Therefore, it is very important to keep the temporal and spatial schematization relatively simple and not to try to include a high level of local detail, because this does not support strategic level planning and might even present results with a false sense of accuracy. This requires trade- offs to be made during the collaborative modeling process between the amount of detail to be included in the models and the strategic purpose for which they will be applied.

(19)

Hydrological models SPHY and Wflow 4

4.1 Spatially distributed hydrological modeling

The hydrological properties and future hydrological changes of single catchments or entire river basins are typically assessed with hydrological models. Hydrological models are simplified representations of components of the hydrological cycle, as shown in Figure 4-1.

Figure 4-1. Overview of the relation between the real world situation and the (conceptual) hydrological model.

Many hydrological models are used and depending on the model’s purpose they are based on different concepts and level of detail included. The simpler hydrological models are empirical models. These models are largely based on observed relationships rather than based on simulated physical processes. Usually they are based on the relationship between precipitation and discharge. These models are often lumped, treating a complete watershed as a homogeneous whole. On the other side are the more complex, physically-based models.

These models have detailed, descriptions of physical processes, and often need a large number of input variables. They can include energy-balance modeling besides water balance modeling. Physically-based models are often distributed, dividing a watershed into elementary units like grid cells and calculating flows between them. There is a large transition zone between the empirical and physically-based models in terms of the detail of representation of physical processes. Models in the transition zone are often referred to as conceptual models. Similarly there is also a transition in spatial discretization between lumped models and distributed models. The models in the transition zone are often categorized as semi-distributed, dividing a watershed in different areas or sub basins. In this project, a fully distributed modeling approach is applied, with two complementing models: SPHY for the upstream mountainous part of the Ganga basin, and WFLOW-SBM for the downstream part of the Ganga basin. The advantage of using a fully distributed approach over a lumped modeling approach is that the spatial variations in physical properties within the basin, as shown in the example in Figure 4-2, can be well represented, and therefore model output for every grid cell can be used, rather than only at a basin’s outlet, as is the case for a lumped approach. Besides, better knowledge about the hydrology in different parts of the basin can be obtained.

(20)

Figure 4-2. Example of the differences in hydrological response of the system within a river basin. This implies the importance of taken into account the distribution of these processes in the hydrological model.

Figure 4-3. Examples of gridded data that can be used as input for the models, or that is generated by the models.

In a distributed approach the basin is divided in grid cells of equal size, and for each grid cell the physical processes that contribute to changes in the grid cell’s water balance are

(21)

simulated. Besides transport of water between grid cells is calculated, thus representing the flow of water from upstream to downstream.

Also the input data (precipitation, evaporation, temperature) for the distributed hydrological models is distributed over the basin (see Figure 4-3). It is not needed to lump the rainfall over the whole (sub)catchment. The distribution of the meteorological inputs comes much closer to what happens in reality, therefore the distributed models represent the real world in much more detail.

4.2 Mountain hydrology with SPHY 4.2.1 Concepts

The Spatial Processes in Hydrology model (SPHY) is a spatially distributed leaky bucket type of model, and is applied on a cell-by-cell basis. The main terrestrial hydrological processes are described in a conceptual way so that changes in storages and fluxes can be assessed adequately over time and space. SPHY is written in the Python programming language using the PCRaster (Karssenberg et al., 2001, 2010; Schmitz et al., 2013) dynamic modeling framework.

Figure 4-4: SPHY modeling concepts. The fluxes in grey are only incorporated when the groundwater module is not used. Abbreviations are explained in the text.

SPHY is grid based and cell values represent averages over a cell. For glaciers, sub-grid variability is taken into account: a cell can be glacier free, partially glaciered, or completely

(22)

with snow or land that is free of snow. Land that is free of snow can consist of vegetation, bare soil, or open water. The dynamic vegetation module accounts for a time-varying fractional vegetation coverage, which affects processes such as interception, effective precipitation, and potential evapotranspiration. Figure 4-4 provides a schematic overview of the SPHY modeling concepts.

The SPHY model provides output variables that can be selected based on the preference of the user. Spatial output can be presented as maps of all the available hydrological processes, i.e., actual evapotranspiration, runoff generation (separated by its components), and groundwater recharge. These maps can be generated on a daily basis, but can also be aggregated at monthly or annual time periods. Time series can be generated for each cell in the study area. Time series often used are stream flow, actual evapotranspiration and recharge to the groundwater. For more detailed description of the concepts of modeling high mountain hydrology in SPHY, please refer to the inception report, the theoretical manual (Terink et al., 2015b) , and journal paper (Terink et al., 2015a).

4.2.2 Set-up and link with WFlow

The SPHY-model is set up for the upstream, mountainous part of the Ganga basin. This domain covers large parts of Himachal Pradesh and Uttarakhand in India, large part of Nepal and parts of China on the Tibetan Plateau. The model extent, which is derived from the hydrologically corrected HydroSheds SRTM DEM (Lehner et al., 2006) is indicated in the model’s projection in Figure 4-5. The discharge generated in the SPHY model domain culminates at the model’s nine outflow locations (Figure 4-5). The simulated discharges at these locations feed into the WFLOW model, which is set up for the downstream parts of the Ganga basin.

Figure 4-5: SPHY model extent (green rectangle), outflow locations (green dots), and their catchments.

(23)

Table 4-1: Properties of SPHY model upstream Ganga.

Projection Asia South Lambert Conformal Conic

(EPSG:102030)

Spatial resolution 1x1 km

Simulated grid cells 260531

Time step 1 day

Most important model inputs are a digital elevation model (DEM), meteorological forcing, glacier extents, soil properties and land use types. Initial model setup is done with data available in the public domain, and can be refined with local data where available.

4.2.3 Data requirements

SPHY needs static input maps as well as series of input maps for meteorological forcing. As static input SPHY needs a digital elevation model (DEM) and a local drain direction map and slope map which can be derived from the DEM. Furthermore it needs a land cover map with associated evapotranspiration coefficients assigned to each land cover type, soil map with quantitative soil properties for the topsoil and subsoil, map of glacier outlines and distinction in debris-covered and debris-free glacier surfaces. As map series input it needs daily grids of precipitation (mm/day), and daily mean air temperature, daily maximum air temperature and daily minimum air temperature (all in ˚C). Detailed information on the input data is provided in Annex 1.

4.2.4 Calibration and results

The calibration strategy for SPHY is based on calibration of simulated discharge to observed discharge. Calibration is done for monthly averaged discharges and focus of the analysis of calibration results is on monthly and annual discharge totals. The model performance is quantified by the Nash-Sutcliffe efficiency (NSE, (Nash and Sutcliffe, 1970)), Pearson’s correlation coefficient (or R²), and bias (or relative volume error (RVE)). A ten year period is used for model calibration, whereas a period of 5 (different) years is used for an independent validation of the model’s performance. Obviously, the locations used for calibration and validation are largely determined by data availability. Ideally, the locations used for calibration and validation represent a large range of catchment types in terms of catchment area, hypsometry, degree of glaciation, climatic regime, soil and land use type.

At least for five gauging stations in Nepal, daily data is available for 1998-2007. This data is property of the Nepal Department of Hydrology and Meteorology (DHM) and can be used for model calibration and validation, but cannot be published. The locations of these stations are indicated in Figure 4-5 with green dots. The station locations from CWC in the upstream part of the Ganga basin are indicated in the same figure with red dots. The most useful of these stations seem to be the stations with ID’s 18, 20, 14, 12 and 9, but their data availability is not yet known. These stations are located near the outlets of the upstream model domain and/or just above a major dam.

(24)

Figure 4-6: Maps of some model inputs at model resolution. Top: digital elevation model, middle: glacier outlines, bottom: land cover types.

(25)

Figure 4-7: Possible gauging station locations identified for model calibration. Station properties are listed in Table 4-2 .

Table 4-2: Locations of gauging stations identified in the upstream Ganga basin. ID’s are indicated in Figure 4-7.

ID StatName StatCode River Lat Lon

1 Badrinath GG2OOV5 Ganga/Alaknanda 30.771 79.494

2 Bausan GYOOOZ3 Yamuna 30.516 77.928

3 Chandrapuri GG25O15 Ganga/Alaknanda/Mandakini 30.438 79.073

4 Deoprayag(a1) GG1OOA1 Ganga/Bhagirathi 30.150 78.598

5 Dharchula Ganga/Mahakali 29.846 80.543

6 Haripur GYXOOD4 Yamuna/Tons 30.526 77.852

7 Jateon Barrage GYWOOK6 Ganga/Yamuna/Giri 30.589 77.484

8 Jauljibi GGU64D1 Ganga/Ghaghra/Sharda/Gauriganga 29.750 80.367

9 Jhulaghat Ganga/Mahakali 29.571 80.383

10 Joshimath GG2OOS3 Ganga/Alaknanda 30.565 79.561

11 Karanprayag GG2OOK2 Ganga/Pinder 30.256 79.221

12 Nandkeshi GG26OJ4 Ganga/Alaknanda/Pinder 30.083 79.508

13 Naugaon GYOOOZ8 Ganga/Yamuna 30.792 78.135

14 Rudraprayag_BC GG2OOG5 Ganga/Alaknanda 30.273 78.961

15 Srinagar GG2OOD5 Ganga/Alaknanda 30.226 78.776

16 Tawaghat GGU65C3 Ganga/Ghaghra/Sharda/Kali/Dhauliganga 29.933 80.580

17 Tehri (Zero Point) GG11OA1 Ganga/Bhagirathi 30.357 78.483

18 Tuini(P) GYX1OA1 Yamuna/Pabbar 30.960 77.854

19 Tuini(T) GYXOOM4 Yamuna/Tons 30.940 77.847

20 Uttarkashi GG1OOK4 Ganga/Bhagirathi 30.739 78.356

21 Yashwant Nagar GYWOOP5 Yamuna/Giri 30.887 77.206

A Turkeghat 604.5 Arun River 27.33 87.18

B Barhbise 610 Bhote Koshi 27.79 85.88

C Pachuwarghat 630 Sunkoshi River 27.56 85.75

D Khurkot 652 Sunkoshi River 27.33 86.00

E Rabuwa Bazar 670 Dudhkoshi River 27.27 86.65

F Mulghat 690 Tamor River 26.93 87.32

G Chatara 695 Saptakoshi River 26.87 87.15

(26)

To aid in calibration the automated “Limited Memory Algorithm for Bound Constrained Optimization” (L-BFGS-B) is used (Byrd et al., 1995). This algorithm is applied to the NSE of the correlation between simulated and observed discharge, to find the parameter set which results in the maximum NSE. One calibrated parameter set will be used for the entire upstream model domain, because not all catchments have gauges and the spatial variability of model parameters cannot be assessed for the entire model domain. Model parameters to be calibrated are listed in Table 4-3.

Table 4-3: SPHY model calibrated parameters.

Parameter name Symbol

Degree day factor for clean ice glaciers DDFCI

Degree day factor for debris covered glaciers DDFDC

Parameter name Symbol

Degree day factor for snow DDFS

Critical temperature for precipitation to fall as snow TCrit Water storage capacity of snow pack SnowSC Minimum slope for gravitational snow transport Sm

Minimum snow holding depth ShdMin

Snow holding depth threshold function parameters SS1 SS2 Potential sublimation function SubPot

Base flow recession constant αGW

Routing recession coefficient kx

4.3 Basin hydrology with WFlow 4.3.1 Concepts

The Wflow-SBM model is, like the SPHY model, a spatially distributed model and is applied on a cell-by-cell basis. The main terrestrial hydrological processes are described in a conceptual way so that changes in storages and fluxes can be assessed adequately over time and space. Wflow is written (just as SPHY) in the Python programming language using the PCRaster (Karssenberg et al., 2001, 2010; Schmitz et al., 2013) dynamic modeling framework.

Wflow is grid based and cell values represent averages over a cell. The following processes are simulated and will be explained in more detail:

§ Snow module

§ Rainfall interception module

§ Soil module

§ Kinematic wave routing module

(27)

Snow module

Precipitation that falls in regions that are very cold, like in high, mountainous areas, can fall as snow. How it is decided whether precipitation falls as snow or rain is calculated in the snow routine. For this, a degree-day factor is used. This routine uses a threshold temperature (tt), below which precipitation in principle will fall as snow. Also in reality, there is no very strict line between precipitation falling as snow and precipitation falling as rain. There is mixed zone where both rain and snow can fall. This is defined in the snow routine as an interval between an upper and a lower temperature (ttint) in which snow and rain can fall at the same time. This is also represented in Figure 4-8.

Figure 4-8. Schematic overview of the snow melt routine using the degree-day factor.

Snow is stored in the model and can build up as long as the temperature is below the threshold temperature (tt). When the temperature comes above the threshold temperature, snow starts to melt according to the degree-day factor (cfmax). This parameter controls how much snow can melt (in mm), per day and per degree Celcius temperature difference between the actual temperature and the threshold temperature. In Figure 4-9 the effect of taking different values for the degree-day factor is shown.

Water can also refreeze when the temperature degreases. This is controlled by the CFR parameter. In the end, water that is melted and does not refreeze is transferred to the soil routine. A complete schematic overview of the snow routine is given in Figure 4-10.

(28)

Figure 4-9. Effect of the degree-day factor (cfmax) controlling the snow melt for a constant temperature difference of 2 degree Celcius.

Figure 4-10. Schematic overview of the snow routine.

(29)

Rainfall interception module

Rainfall interception is an important process in the hydrological cycle. Depending on the density of the canopy, more or less water can be stored. From the canopy, evaporation takes place. This water will therefore not come to runoff. If this process is ignored, more water enters the runoff process in the model, generating more runoff.

The rainfall interception is controlled by a set of parameters, which are mainly based on the land-use / land-cover in the basin. Dense forest stores more water than, for example, open grass-land. By linking the parameter values to the different classes of land-use, these differences are taken into account in the Wflow model. The processes important for interception are schematically presented in Figure 4-11.

Figure 4-11. Schematic overview of the rainfall interception module.

Soil module

The soil module in the Wflow-SBM model is represented as a single bucket model. This is schematically represented in Figure 4-12. Within this bucket, the water table can go up and down, depending on the sum of the in- and outflows in the bucket. The movement of the water through the soil is controlled by a set of parameters, setting the properties of the soil.

These parameters include the maximum depth of the different zones (i.e. the saturated and un-saturated zones), the porosity of the soil controlling the flow velocity through the soil (both vertically and horizontally) and the rooting depth of the vegetation controlling the amount of transpiration through the vegetation.

(30)

Water enters the soil module from the snow module, the interception module and from direct rainfall. There is also linkage with the kinematic wave module. Water moves from the soil module to the kinematic wave module, but water can also infiltrate through the river bottom and enter the soil. This might be of special interest in the Ganga basin, where it is known that the groundwater table in some areas can be (far) below the river bottom.

Figure 4-12. Schematic overview of the soil module.

One of the most important parameters for calibration is the M parameter. This parameter controls the hydraulic conductivity. Using the M parameter, non-linear decrease of the hydraulic conductivity with depth is introduced. This represent the fact that deeper soils are generally more compacted, resulting in lower hydraulic conductivity.

For higher values of the M parameter, the hydraulic conductivity at larger depth is higher, meaning that more interaction between the saturated and unsaturated part of the soil is possible. It is very difficult to determine the value forM. Therefore, the value forM is normally determined by calibration of the model. In Figure 4-13 the relation between hydraulic conductivity and depth is shown for different values of theM parameter.

(31)

Figure 4-13. Relation between hydraulic conductivity (K in mm/day) and depth (zi in mm) for different values of the M parameter.

Kinematic wave module

The water in the Wflow model is routed through the river network using a kinematic wave routine. The kinematic wave is an approximation for one-dimensional dynamic waves in the river. The kinematic wave model solves a simple form of the shallow water equations, in which two terms (i.e. the inertia and the pressure-differential terms) are assumed to be insignificant and are thus ignored. It basically comes down to solving the continuity equation, for which simple formulae like Manning or Chezy formulae can be used (Miller, 1984).

Different roughness values can be set for different stretches of the river, controlling speed in which the water flows through the river system.

4.3.2 Set-up and link with RIBASIM

For the building of the Wflow model, the most important data source is the elevation data. For the project the SRTM90 dataset is used (Jarvis et al., 2008). Since the resolution of the 90 meter SRTM Digital Elevation Model (DEM) is too high for using it in the model, the DEM is resampled to a 1000*1000 meter resolution for the whole Ganga basin. From the DEM, the local drainage direction map (LDD) is derived. This map contains the information about all possible direction the water can flow to. From the LDD the river map is abstracted. In Figure 4-14 the Ganga basin extend is shown, including the elevation data (DEM) and river network derived from the DEM.

(32)

Figure 4-14. Overview of the Ganga river basin upstream from Farraka dam, with the DEM and rivers derived from the DEM (elevation in meters).

As described in section 4.3.1, most hydrological processes depend largely on the land use.

Therefore, a map of the different land use classes is needed to distinguish the different hydrological processes in the Ganga river basin. In Figure 4-15 an example of the ESA GlobCover map for the Ganga basin is shown. The map clearly indicates the differences in land use and land cover over de Ganga basin. This provides crucial information to build the hydrological models.

The soil properties in the Ganga basin can be derived from the Digital Soil Map of the World, as shown in Figure 4-16. The soil properties can be related to the hydrological processes and parameters. Different soil types will react different, hence different parameterization based on the soil types will increase the realism of the model.

(33)

Figure 4-15. Overview of the land use classes within the Ganga river basin, based on the ESA GlobCover dataset.

The land classes have been resampled from the original classes for better use in the hydrological modeling.

A first model was setup during the course, given in March 2016. This model was completely based on the datasets found in the public domain. Some properties of the model are given in Table 4-4. The next steps of setting up the model include:

§ Refine the model based on the collected data.

§ Calibrate and validate the model.

§ Connect the model to the RIBASIM model (Chapter 4).

§ Run the model for the selected period and (climate) scenarios.

Table 4-4: Properties of Wflow model of the complete Ganga basin upstream Farraka dam.

Projection Asia South Lambert Conformal Conic (EPSG:102030) Spatial resolution 1x1 km

Simulated grid cells 2,563,074 Time step 1 day

(34)

Figure 4-16. Overview of the soil type classes within the Ganga river basin, based on the Digital Soil Map of the World. The soil type classes have been resampled from the original classes for better use in the hydrological modeling, based on their dominant soil type.

To connect the Wflow model to the RIBASIM model, the output of the Wflow model is used as input for the RIBASIM model at selected locations. Since the Wflow model simulates the natural flow, large infrastructural elements in the system, like dams and reservoirs, will not be included in the model in much detail. This is typically the domain of the RIBASIM model.

Therefore, logical connection points are the inflows to large reservoirs and dams. Wflow will calculate the (natural) inflow and RIBASIM will route this water through the system downstream.

4.3.3 Data requirements

Wflow needs static input maps for the schematization of the model as well as series of input maps for meteorological forcing of the model. As static input Wflow needs a digital elevation model (DEM) and a river network to “train” the derivation of the local drainage direction (LDD) map. Furthermore Wflow needs land cover map soil maps, which can be linked to the dominant hydrological processes. Information about the phenology (i.e. the temporal change of the canopy thickness) is mandatory, but can increase the quality of the model.

As map series input Wflow needs daily grids of precipitation (mm/day), daily mean air temperature (in ˚C) and preferably, potential evaporation (in mm/day). Also meteorological station data is required to validate the gridded datasets and, if needed, to correct the gridded

(35)

datasets for e.g. elevation. Most of these datasets can be obtained from the public domain. If more detailed data is available locally, these data can be used to refine the Wflow model.

For the calibration and validation of the Wflow model (see next section), discharge data is highly relevant. Ideally, the discharge data is provided with a daily frequency, since the model will be run with this time step. However, if daily data will not be available, monthly average (or total) discharges also suffice for a global optimization of the Wflow model. Detailed information on the input data is provided in Annex 1.

4.3.4 Calibration and results

The Wflow model is calibrated based on observed meteorological and discharge measurements. For this, the requested data must be obtained from CWC, see Annex 1. The overall performance of the model is analyzed based on the monthly average (or total) flow.

For a check on the total volume of water, also annual total discharge will be checked.

The comparison of the observed and simulated discharges is done at river locations where the flow is mainly undisturbed. Typically, the Wflow model is calibrated on stations which are located upstream of (large) infrastructural changes in the river. A stepwise approach for calibration is described below:

1) Select relevant locations with good observations for calibration:

a) Based on geographical location.

b) Data availability.

c) Data quality.

d) Period for which the data is available.

2) Do a large number of simulations with different parameter values.

3) Analyze the different simulation results with observations:

a) Based on daily hydrograph comparison (if available).

b) Based on mean monthly discharges (or discharge regime).

c) Based on annual total volumes.

4) Calculate for each simulation the value of the performance measures, like Nash-Sutcliffe efficiency (NSE, (Nash and Sutcliffe, 1970)), Pearson’s correlation coefficient (or R²) and bias (or relative volume error (RVE)).

5) Select the parameters for which the model performs best, defined by a scored index of the different performance measures.

Validate the model by running the model for a different period.

(36)

Groundwater flow modeling 5

The chapter 9 the set-up of the SW-GW assessment will be described. An important contribution to that assessment are the analytical results based on the River Basin Model tool, especially the part simulation the SW-GW interaction; iMOD. This chapter describes the set-up of this groundwater flow model in relation to the River Basin Model tool.

5.1 Concepts in MODFLOW

Deltares developed the iMOD software package (Vermeulen 2016) which will be used to support the analytical work of understanding the Ganga Basin groundwater system. iMOD is an easy to use Graphical User Interface combined with an accelerated Deltares-version of MODFLOW (McDonald 1988) with fast, flexible and consistent sub-domain modeling techniques. iMOD facilitates very large, high resolution MODFLOW groundwater modeling and also geo-editing of the subsurface and is very powerful in the visualization of model data from different sources. iMOD is open source since June 2014.

MODFLOW is the U.S. Geological Survey flow model for the simulation of flow of groundwater through aquifers. It is a finite-difference model and provides for different modules, each modeling a specific phenomenon.

Recharge module

MODFLOW provides for several modules to calculate the groundwater system, depending on the processes simulated. Recharge can be calculated by Modflow using modules like EVT or UZF. However, because SPHY and WFlow calculate the spatial net recharge component, the groundwater model will use this variable as input to the ground water model through the recharge module RCH.

River module

Rivers are element that can gain or lose water, depending on the surface water level and groundwater head. The RIV module of MODFLOW provides for this process. The resistance to flow between the compartments groundwater and surface water is the lumped parameter Conductance [m²/d]. In standard Modflow, the river line elements are gridded to the model scale. In iMOD rivers are represented in ISG format. The ISG-file format is developed to capture all relevant information used by surface water elements in direct relation with groundwater. It water level, bottom level, infiltration factor, conductance/resistance, and moreover, the actual outline of the surface water element.

Drainage module

Streams without managed water levels, the fixed Rivers are elements that can gain or lose water, depending on the surface water level and

Well module

For agriculture, industry and domestic use, groundwater is pumped from shallow or deep aquifers. In Modflow the WELL module contains the point location, depth and volume pumped. The pumping rate in some cases will have a seasonal variation that can be modeled with this module based on input from RIBASIM.

(37)

5.2 Set-up

For both the groundwater model and the hydrological model, the model domain is chosen the same in order to organize a smooth model coupling. While the active domain for the groundwater model is smaller than for the hydrological models, part of the groundwater model is defined inactive.

For this project the “Asia_South_Lambert_Conformal_Conic” projection is used (also known as EPSG: 102030). In that projection the window to be modeled is set to the following coordinates:

- XY lower left: -6.087.000 / 3.599.000 - XY upper right: -4.257.000/ 4.899.000 - X distance: 1830 km / Y distance: 1300 km

Because the initial model cell size is 1 km x 1km the number of columns and rows is respectively 1830 and 1300. The number of model layers depends on the hydrogeological characteristics of the subsoil. In aquifers the groundwater flow is in general horizontal while the flow in aquitards is vertical. The vertical extent (depth) of the model depends on the depth to where the influence of the measures and scenario’s reaches. An impermeable layer usually is the boundary of a model.

In a pre-processing phase of a model run, all model data as described in paragraph 0 is rescaled to the actual model cell size. During the model development, attention will be paid to the process of rescaling river elements (lines) to a model cell of 1000x1000 meter. Processes that are relevant at small scales might become irrelevant at larger scales.

The initial model scale is 1x1 km. In case the conclusion is that it is to coarse, iMOD can (locally) zoom in to a finer scale, for instance 500x500 m.

The focus of the sw-gw analysis is on the most important Groundwater Management Units as explained in chapter 9.3.

5.3 Initial model results

Together with the Faculty of Geoscience (University of Utrecht) Deltares worked on the development of a global scale water demand model PCR-GLOBWB (for more detail, see http://pcraster.geo.uu.nl/projects/applications/pcrglobwb). PCR-GLOBWB is a global hydrology and water resources model. It is built to simulate global terrestrial hydrology and human water use at daily time step and 5 arcminute resolution (approximately 10 km at the equator).

The initial groundwater model for the Ganga basin is a cut out of the existing global PCR- GLOBWB model. During the first phases of the project, the groundwater model will be detailed, whenever new datasets are available. Figure 5-1 shows the average piezometric heads in the first aquifer calculated with this initial model.

(38)

Figure 5-1: Calculated head [m] in the Ganga Basin with a test version of the Ganga groundwater model

5.4 Links with WFlow and RIBASIM

The groundwater model is loosely coupled with both WFlow and Ribasim. The definition of a loosely coupling is: two (or more) individual models are coupled via the exchange of model results. The output of one model forms the input of the other. This paragraph describes the input and output on which the coupling is based.

Figure 5-2: Schematic representation of technical interaction between Ribasim, WFlow and iMOD

(39)

Modflow-WFlow

WFlow calculates surface run off and evaporation. This provides information on the volume of water that recharges the groundwater. This data is input for the Modflow model. While both Modflow and WFlow are distributed models, the data exchange is on cell basis.

Modflow-Ribasim

RIBASIM describes the management and use of water. Water can be used from rivers and canals or from groundwater (abstraction wells). The output of the RIBASIM model is a demand for a volume of groundwater. This output is input for Modflow as is defines the abstraction rates in the WEL module.

Not all well numbers and well locations are known. In those cases an artificial well distribution has to be developed based on expert judgment. An example is shown in Figure 5-3.

Modflow is a distributed model while Ribasim has a lumped set-up so the data exchange is based on ID numbers of the Ribasim groundwater reservoirs.

Modflow calculates the dynamic of the groundwater level. In case the groundwater level drops it can get out of reach of an abstraction well. In that case it will reduce the abstraction rate, even to zero. It might be necessary to use the groundwater level as input to check whether Ribasim is allowed to claim a volume of groundwater.

Another result of a Ribasim calculation is the irrigation loss to the groundwater compartment.

This flux is input for the groundwater model. Ribasim also calculates the remaining flux (water level) in rivers. This water level is transferred to the groundwater model as a boundary condition. With this new boundary conditions, iMOD calculates the fluxes to (gaining) and from (loosing) the river system. This output is redirected to the Ribasim model and is used for a rerun of Ribasim. In this way an iteration process is created to ensure consistent results from both models.

(40)

Figure 5-3: Example of artificial distribution of lumped groundwater wells over the Ganga basin, one for each Taluka area (blue: Ganga Basin).

5.5 Calibration process

Calibration of the groundwater model is the process of creating a model that optimally represents the geohydrological phenomena needed for the purpose given. The users must have confidence in the model's predictions.

Two types of measurements are used for the calibration of the groundwater model:

piezometric head measurements and surface water discharge measurements. The technic iMOD provides for parameter optimization of a groundwater model is a package called PST which is based on an existing optimization code PEST (Doherty 2010).

The calibration process distinguishes 4 phases:

- Selection and analyzing of the right measurements;

- Sensitivity analysis of model parameters followed by model (concept) optimization - Final parameter calibration

- Validation Data selection

The comparison of the observed and simulated data is done based on piezometric head measurements and measurements of surface water discharge.

Discharge measurements are selected at river locations where the flow is mainly undisturbed.

These are typically stations which are located upstream of (large) infrastructural changes in

(41)

the river. The station locations will be selected in consultation with the team developing WFlow.

A large number of groundwater measurements are available from the CGWB and other sources with different quality, measuring method, frequency and from specific locations (near river, near road, near abstraction).

To select relevant locations with reliable and useful observations for calibration the selection will be based on (Barthel 2016):

- geographical location (e.g. measurements distributed over all relevant model domains such as sub catchments);

- data availability (preferable long and continuous time series);

- data quality (check on outliers, strange drift or jumps within series);

In order to collect field data about surface water groundwater interaction, special attention is paid to piezometer locations near river gauging stations (Brownbill 2011).

The model will run with a time step of one month. This means that in a special process all data for calibration is transformed into monthly based time series.

Sensitivity analysis

Sensitivity analysis provides insight in those parameters that influence the outcome of the model most. In case the first model results and measurements differ more than accepted, the reliability of these parameters will be tested and if necessary, more / detailed values will be necessary.

The conclusion from the analysis can also be that mistakes were made in compiling the model parameter set or that an important model concept is missing. In this process of examining the results, a final model for calibration is defined.

Calibration

The period 1985 to 2000 is proposed for calibration of the models and 2001 to 2015 for validation of the models. The calibration period includes one major El Niño Southern Oscillation (ENSO) event (1997-98) and the validation period another one (2015). ENSO events are for India associated with relatively little precipitation in the monsoon period.

The time step of the models will be one month. Therefore, calibration and validation will use average monthly values. The water resources analysis will cover the whole period for which meteorological forcing functions can be obtained: 1901-2015 if validation shows that the data are reliable for this whole period.

The calibration not only focusses on the groundwater model, it will be an integrated process together with the models WFlow and Ribasim. Figure 5-4 gives an impression of these relations.

The DEMAND for irrigation and industrial /domestic users is calculated by Ribasim. The effect of this water demand on the surface water levels (h) in both rivers and canals is calculated by WFlow. Modflow uses new rivers stages (WFlow), the abstraction of groundwater and return flows from irrigation (Ribasim) to calculate both the groundwater head as well as fluxes to or from the river and canals. These fluxes are input for WFlow and Ribasim.

(42)

Figure 5-4: Schematic relations between hydrological models and groundwater model

Validation

Validation is done by comparing the model results with a set of monitoring time series that was not part of the calibration process. That can be a small set from all monitoring locations.

The other option is to select the monitoring results for a complete year. For the River Basin Model the years 2006 and 2015 are proposed for validation of the models.

5.6 Data requirements

Important step is to understand the groundwater system in order to understand the model results. For the understanding of the system the following maps are necessary:

- Water loss and water gain Map of the Ganges river and the distributaries (based on existing studies)

- Basin wide hydraulic head map (based on studies, models etc. incl. drawdown map) - Simple hydrogeological map (determine homogenous areas based on hydrogeological

sequences), for example developed by MacDonald (2015).Depth to brackish – salt groundwater map (mapping the existing fresh groundwater body)

- Initial groundwater quality assessment (basin wide map and/or strategic transects) - Develop a basin wide groundwater extraction map (based on: (1) existing pumping well

locations and extraction rates, (2) irrigated land use (type, number of harvests), estimated on evapotranspiration demand, (3) (estimated) urban groundwater extractions)

- Data on groundwater extraction rates and locations: (1) Drinking water, (2) industry, (3) agriculture

- Subsidence map or information

For the development of the numerical groundwater model data requirements are shared with other activities like the hydrology modeling. It applies e.g. for the Digital Elevation Model, land use, precipitation and evaporation, surface water system (river, canal, drainage) including its characteristics (bed level / width / depth).

The specific data requirements for the groundwater model are:

- Characteristics of the aquifer system (depth, hydraulic conductivity, yield);

- Information on Faults (map or information about horizontal hydraulic resistivity);

- Groundwater extraction: spatial distributed data on rates over time, well depth and user group. An example is given in Table 5-1. (calibration)

- Time series of hydraulic heads (a selection of available time series for calibration)

(43)

The process of data gathering started with the easy available data in order to develop an initial groundwater model. During the project data requirements might be detailed both in time and space. This approach prevents for over focusing on data gathering.

Table 5-1: Estimated water availability and water use (source: CGWB 2014).

Basin Planning for Ganga River Basin in India

Conceptualization of River Basin Model, Surface water - Ground water Interaction Analysis, and Environmental Flow Assessment

2 December 2016

Basin Planning for Ganga River Basin in India

Contents

Introduction 1

A - RIVER BASIN MODEL CONCEPTUALIZATION Context: The collaborative modeling process 2

Components of the River Basin Model and their interaction 3

Hydrological models SPHY and Wflow 4

Groundwater flow modeling 5